[PYTHON/KERAS] 학습 모델 저장하기

■ 학습 모델을 저장하는 방법을 보여준다.

 

▶ 예제 코드 (PY)

import keras.datasets.mnist as mnist
import keras.layers as layers
import keras.models as models
import keras.utils.np_utils as np_utils
import numpy as np

np.random.seed(3)

# 데이터를 로드한다.
(trainInputNDArray, trainCorrectOutputNDArray), (testInputNDArray, testCorrectOutputNDArray) = mnist.load_data()

# 훈련 데이터에서 검증 데이터와 훈련 데이터를 나눈다.
validationInputNDArray         = trainInputNDArray[50000:]
validationCorrectOutputNDArray = trainCorrectOutputNDArray[50000:]
trainInputNDArray              = trainInputNDArray[:50000]
trainCorrectOutputNDArray      = trainCorrectOutputNDArray[:50000]

# 데이터셋 전처리
trainInputNDArray      = trainInputNDArray.reshape(50000, 784).astype("float32") / 255.0
validationInputNDArray = validationInputNDArray.reshape(10000, 784).astype("float32") / 255.0
testInputNDArray       = testInputNDArray.reshape(10000, 784).astype("float32") / 255.0

# 훈련/검증 데이터를 랜덤하게 섞는다.
randomTrainIndexNDArray = np.random.choice(50000, 700)

trainInputNDArray         = trainInputNDArray[randomTrainIndexNDArray]
trainCorrectOutputNDArray = trainCorrectOutputNDArray[randomTrainIndexNDArray]

randomValidationIndexNDArray = np.random.choice(10000, 300)

validationInputNDArray         = validationInputNDArray[randomValidationIndexNDArray]
validationCorrectOutputNDArray = validationCorrectOutputNDArray[randomValidationIndexNDArray]

# 정답 데이터를 인코딩 한다.
trainCorrectOutputNDArray      = np_utils.to_categorical(trainCorrectOutputNDArray)
validationCorrectOutputNDArray = np_utils.to_categorical(validationCorrectOutputNDArray)
testCorrectOutputNDArray       = np_utils.to_categorical(testCorrectOutputNDArray)

# 모델을 정의한다.
model = models.Sequential()

model.add(layers.Dense(units = 64, input_dim = 784, activation = "relu"))
model.add(layers.Dense(units = 10, activation = "softmax"))

model.compile(loss = "categorical_crossentropy", optimizer = "sgd", metrics = ["accuracy"])

# 모델을 학습시킨다.
model.fit(trainInputNDArray, trainCorrectOutputNDArray, epochs = 1000, batch_size = 100, validation_data = (validationInputNDArray, validationCorrectOutputNDArray))

# 학습 과정을 조회한다.
evaluateList = model.evaluate(testInputNDArray, testCorrectOutputNDArray, batch_size = 32)

print("")
print("loss     : " + str(evaluateList[0]))
print("accuracy : " + str(evaluateList[1]))

# 모델을 저장한다.
model.save("model.data")